https://github.com/internetarchive/warcprox/issues/9 record warcprox version in warcinfo metadata, and add --version command line option

setup.py

@@ -5,6 +5,26 @@ from setuptools.command.test import test as TestCommand
 import sys
 import setuptools 
 
+VERSION_BYTES = b'1.3'
+
+def full_version_bytes():
+    import subprocess, time
+    try:
+        commit_bytes = subprocess.check_output(['git', 'log', '-1', '--pretty=format:%h'])
+
+        t_bytes = subprocess.check_output(['git', 'log', '-1', '--pretty=format:%ct'])
+        t = int(t_bytes.strip().decode('utf-8'))
+        tm = time.gmtime(t)
+        timestamp_utc = time.strftime("%Y%m%d%H%M%S", time.gmtime(t))
+        return VERSION_BYTES + b'-' + timestamp_utc.encode('utf-8') + b'-' + commit_bytes.strip()
+    except subprocess.CalledProcessError:
+        return VERSION_BYTES
+
+version_bytes = full_version_bytes()
+with open('warcprox/version.txt', 'wb') as out:
+    out.write(version_bytes)
+    out.write(b'\n');
+
 # special class needs to be added to support the pytest written dump-anydbm tests
 class PyTest(TestCommand):
     def finalize_options(self):
@@ -17,9 +37,8 @@ class PyTest(TestCommand):
         errno = pytest.main(self.test_args)
         sys.exit(errno)
 
-
 setuptools.setup(name='warcprox',
-        version='1.2',
+        version=version_bytes.decode('utf-8'),
         description='WARC writing MITM HTTP/S proxy',
         url='https://github.com/internetarchive/warcprox',
         author='Noah Levitt',
@@ -27,6 +46,7 @@ setuptools.setup(name='warcprox',
         long_description=open('README.rst').read(),
         license='GPL',
         packages=['warcprox'],
+        package_data={'warcprox':['version.txt']},
         install_requires=['pyopenssl', 'warctools>=4.8.3'],  # gdbm not in pip :(
         dependency_links=['git+https://github.com/internetarchive/warctools.git#egg=warctools-4.8.3'],
         tests_require=['requests>=2.0.1', 'pytest'],  # >=2.0.1 for https://github.com/kennethreitz/requests/pull/1636

warcprox/warcprox.py

@@ -76,6 +76,14 @@ import base64
 import json
 import traceback
 
+def _read_version_bytes():
+    version_txt = os.path.sep.join(__file__.split(os.path.sep)[:-1] + ['version.txt'])
+    with open(version_txt, 'rb') as fin:
+        return fin.read().strip()
+
+version_bytes = _read_version_bytes()
+version_str = version_bytes.strip().decode('utf-8')
+
 class CertificateAuthority(object):
     logger = logging.getLogger('warcprox.CertificateAuthority')
 
@@ -875,7 +883,7 @@ class WarcWriterThread(threading.Thread):
         headers.append((warctools.WarcRecord.DATE, warc_record_date))
 
         warcinfo_fields = []
-        warcinfo_fields.append(b'software: warcprox.py https://github.com/internetarchive/warcprox')
+        warcinfo_fields.append(b'software: warcprox ' + version_bytes)
         hostname = socket.gethostname()
         warcinfo_fields.append('hostname: {}'.format(hostname).encode('latin1'))
         warcinfo_fields.append('ip: {0}'.format(socket.gethostbyname(hostname)).encode('latin1'))
@@ -1175,6 +1183,8 @@ def _build_arg_parser(prog=os.path.basename(sys.argv[0])):
     arg_parser.add_argument('--playback-index-db-file', dest='playback_index_db_file',
             default='./warcprox-playback-index.db',
             help='playback index database file (only used if --playback-port is specified)')
+    arg_parser.add_argument('--version', action='version',
+            version="warcprox {}".format(version_str))
     arg_parser.add_argument('-v', '--verbose', dest='verbose', action='store_true')
     arg_parser.add_argument('-q', '--quiet', dest='quiet', action='store_true')
     # [--ispartof=warcinfo ispartof]